Efficacy of Screening in Women at High Risk: Dutch MRI Screening Study Update

More data have emerged to suggest that MRI is more sensitive than mammography, though less specific, in screening women at high risk for breast cancer occurrence. Moreover, annual screening picks up a high number of ductal carcinomas in situ (DCIS), but also identifies tumors that are already large at diagnosis among BRCA1 mutation carriers.

Jan G.M. Klijn, MD, PhD
Photo Courtesy SABCS/Todd Buchanan 2007

The data come from the second analysis of the Dutch MRI Screening Study (MRISC), presented at SABCS by Jan G.M. Klijn, MD, PhD, professor of medical oncology at Erasmus University and chairman of the Rotterdam Family Cancer Clinic in Rotterdam, the Netherlands. The principal author of the study was A.J. Rijnsburger, MD. The Dutch screening study included women with a cumulative lifetime risk of breast cancer of >=15%, based on genetic or familial risk, but no previous breast cancer. Women were screened every 6 months with a clinical breast examination and once a year by mammography and MRI between 1999 and 2006. Participants were divided into three subgroups according to their estimated cumulative lifetime risk: BRCA1/2 mutation carriers (50% to 85%), a high-risk group (30% to 50%), and a moderate-risk group (15% to 30%). Investigators analyzed the ability of the various screening methods to detect cancers, and also evaluated the characteristics of the cancers that were detected in each risk group. At a median follow-up time of 5.3 years, the screening population included 2177 women with a mean age of 40 years at entry, 605 of whom (28%) were mutation carriers. Ninety-four malignant breast tumors were detected; 78 were observed at screening, and 16 were interval cancers. This included 74 invasive cancers, 19 DCIS, and one of unknown type.

The overall rate of detection was 43 per 1000 women, with the highest rate in BRCA2 mutation carriers (92 per 1000 women). The rate of detection was 70 per 1000 in BRCA1 carriers, 29 per 1000 in the high-risk group and 33 per 1000 in the moderate-risk group, Dr. Klijn said.

The sensitivity, specificity, and positive predictive values (PPV) of the screening modalities are shown in the table.

For 68 of the 94 tumors, test results for both MRI and mammography were available, yielding results for 6626 screening tests. This analysis showed that 40% of tumors were detected only by MRI screening, 18% were detected only by mammography, 26% were detected by both modalities, 3% were detected only by clinical examination, and 13% were interval cancers not detected by screening. The sensitivity, specificity, and positive predictive values (PPV) of the screening modalities are shown in the table. MRI was more sensitive than mammography in detecting invasive cancers (79% vs 38%) but was less sensitive in detecting DCIS (33% vs 67%). "Mammography, therefore, remains a valuable adjunct to MRI," Klijn said. Tumor characteristics differed according to risk group, he added. The moderate-risk group (non-BRCA carriers) had the highest rate of DCIS (37%), the highest proportion of tumors BRCA1 carriers tended to have less DCIS than BRCA2 carriers (7% vs 37%). BRCA1 carriers tended to have larger tumors. In BRCA1 carriers, only 27% of tumors were < 1 cm, compared to 70% in BRCA2 carriers. BRCA1 carriers also had strikingly more interval cancers (37% vs 13%), but node positivity was similar for these two groups. Dr. Klijn said that the high incidence of DCIS is striking, particularly in the BRCA2 mutation carriers. "In the BRCA1 mutation carriers, the high number of interval cancers, low incidence of DCIS, and an unfavorable tumor size at diagnosis are clinically very relevant. Adaptation of the screening schedule may be necessary for this specific subgroup," he said.

Disclosures:

The author(s) have no significant financial interest or other relationship with the manufacturers of any products or providers of any service mentioned in this article.